Abstract: Meteorological satellites have provided useful information for improving weather forecasting, environmental monitoring, and short-term climate prediction. In the field of the weather forecast, satellites provide a powerful means for the forecast of typhoons, rainstorms, hail, sandstorms, and other severe weather conditions. In this study, the microstructure of hail clouds was analyzed by satellite observation data based on nearly a decade of hail event records of Shaanxi, Shandong, Guizhou, and Xinjiang. The comparison between the hail cloud and deep convective precipitation cloud characteristics retrieved by polar orbit satellites showed different cloud properties such as cloud top temperature/effective radius and cloud glaciation temperatures. Based on distinct cloud properties between hail clouds and convective clouds, this work summarized the characteristics and further applied them to the FY-4A geostationary satellite, which captures the hail cycle, which occurred on August 16, 2019, in the Shandong area. Results showed that the satellite has the potential to capture a hail cloud during its developing stage and use it as an application of early warning. The hail cloud shows the following characteristics: (1) There are considerable differences in the cloud’s physical characteristics between hail clouds and deep convective precipitation clouds. Microphysical characteristics of hail clouds observed by satellites are shown in three aspects: Glaciation temperature（T_g）is cooler with an average value of −33°C. The hail cloud reaches the glaciation temperature with a smaller effective radius (<40 μm) with an average of 36.9 μm when the clouds are fully glaciated. It also shows that the smaller the r_eg (effective radius corresponding to glaciation temperature), the stronger the hail cloud. Additionally, hail cloud tops often have a reduction zone of r_e with increasing height. (2) All the studied areas have consistent cloud properties such as a lower T_g, smaller r_eg, and a decreased r_e compared to those of adjacent convective clouds. However, it still showed regional variabilities that indicate the need to establish different indicators for identifying hail clouds for early warning purposes. (3) The case study of the FY-4A geostationary satellite shows that the geostationary satellite can track the evolution of hail clouds. By tracking the hail cloud, the geostationary satellite has a response consistent with that of the polar orbit satellite, providing a method for monitoring and early warning service of hail weather. The geostationary satellite can be used to track the development and evolution of the cloud cluster at any time when the satellite detects a strong hail signal because of the high time resolution. Combining the satellite’s early warning with radar observation, the location of hail occurrence can be determined precisely. (4) Combining the indicators summarized by polar orbit satellites with the FY-4A to track the hail cloud evolution. Four hail storms that occurred in Shaanxi and Shandong were applied for early warnings. Ground observations reported 24 hail events in the two regions, of which the satellite successfully warned 22 times in advance and missed two times. The average early warning time is about two hours before the hail disaster. All of these suggest that the automatic warning of hail by the FY-4A satellite has important practical significance for timely and effective organization and implementation of operational hail mitigation.